Re: Teaching physics to biology students

In article <muMSf.155944$B94.149819@pd7tw3no>, kmuldrew@xxxxxxxxxxx (Ken Muldrew) writes:
Timo Nieminen <uqtniemi@xxxxxxxxxxxxxxxxx> wrote:

On Sat, 18 Mar 2006, Ken Muldrew wrote:

Timo Nieminen <uqtniemi@xxxxxxxxxxxxxxxxx> wrote:

On Fri, 17 Mar 2006, Ken Muldrew wrote:

mmeron@xxxxxxxxxxxxxxxxxx wrote:

I'm sure you're familiar with the Golden Rule, i.e. "he who has the
gold, makes the rules". Those who fund your research are in position
to dictate conditions and in the long run the temptation to do so is
to great to resist.

Yeah, if only they didn't use the metric of paper counting to evaluate
the success of their "initiatives" (My god, what progress we're making
in the fight on cancer!).

There are moves away from paper-counting. We're told it's coming in 2
ways: some journals will be "Tier 1", supposedly the top 20% of them, and
papers therein will count as worth 5 times as much. This appears a bit
crude, but it's a start.

I dont' think so. In fact, I *strongly* disagree.

Ah, but consider what it replaces: counting number of papers.

We just had a power failure and I lost the reply that I was working on
(I'll never recapture the eloquence of that powerful missive; this
poor and hastily scratched bit of dross will have to do in its place).

This may be a difference between medicine and physics (or maybe not, I
really can't say). In medicine we have a cohort of senior people who
are genuinely more interested in serving as elder statesmen (so to
speak) than in continuing to do research. They are like another tier
in the undergrad/grad/post-doc/asst., assoc., full prof cascade. Their
interests are generally more broad than those of an active researcher
(who is by necessity something of a specialist) and their knowledge of
not only the literature, but also of the abilities and talents of many
of the players in several fields, is first rate. I think these people
would genuinely find satisfaction from performing the hard work of
evaluating the work of active researchers. Naturally they would bring
some of their old prejudices to the task, but consider the
alternative: counting number of papers. ;-)

good point:-)

It reminds me of
zero-tolerance policies (any zero-tolerance policies really, you can
pick your favorite). This business of subtracting intelligence from
the process...it's just so ridiculous.

But it's got to be doable by beancounters!

Ah, here's where I haven't made my point. It can *not* be done by any
but those who are skilled in the art. Beancounters have no place in
any reasonable scheme of evaluating scientific productivity.

Of course the intelligence
needs to be subtracted from the process. The alternative would be to hire
people who could think - but would managers tolerate underlings who can
_think_?

I'm dreaming, of course, but I'm dreaming of managers who can _think_
as well. Perhaps the extent of my delusions can inform your homage to
spirits page, if nothing else. ;-)

Oh, many managers can think. What they've problem with is underlings
who can think better. Human creativity doesn't fit well within the
standard schemes of organizing labor, where the basic concept is
"those who can think, manage, the rest work". Individuals who have
brains but are sincerely disinterested in managing others (because, in
their opinion, they've better things to do) don't fit well anywhere in
such system.

Right now there are lots of
smart people spending huge amounts of their time trying to come up
with targeted funding intiatives, and evaluating grant applications
for quality. What a terrible waste! There is not a chance in hell that
they can predict the future well enough to decide who should get
funded (among the reasonable applications; obviously there are dogs
among the pile). What they can do, and do well, is evaluate what
people have done in the past. It isn't easy; one has to spend a lot of
time and effort to place someone's work into the context of their
field and judge the impact. But it can be done (and it has nothing to
do with paper counting). So why don't we do that? Instead of convening
panels of scientists to rank grant applications, get them instead to
rank the impact of the applicant's recent work, and dole out funding
on that basis (of course you can still ask for a proposal, budget,
timeline, etc., it's just that you base your decision on something
that might be relevant to the reasons for providing money in the first
place).

The easiest way to get funded is to promise to do more of the same. It's
hard to get real innovation funded - why, you'd be doing something that
deviates from your past experience, and results won't be guaranteed. How
terribly risky!

In fact, the senior people in medicine who manage to get continuous
funding have an even better scam (here in Canada, health grants are
2-3 times larger than basic science grants, but the success rate is
only around 25%, so one is by no means ever "in" the system on the
basis of past success). They hold back publication of 1-2 years of
work and write their grant proposals based on work that's already done
(with pilot data that is oh-so-convincing (it damn well better be!))
-- this is easier to do if you have 10-12 students and post-docs
churning out data. So successful grants aren't just for more of the
same, they're for work that's guaranteed to be successful (for the
very good reason that it's already been done...and was a success).
Innovation is treated with utter contempt.

Yep.

Allow me to provide a personal anecdote (not so much in support of my
thesis, but certainly in support of my bad attitude). The Canadian
Institutes of Health Research was running a special competition for
"high risk, high reward" proposals. I put in a grant proposing to look
for evidence of communication between cells using light signals (not
as crazy as it sounds; all cells give off a low level of
chemiluminescent light, and there are situations where neural and
muscular electrical signals to excitable cells could cause unwanted
cross-talk, and natural selection tends to use whatever comes to hand,
and a bunch of other reasons that aren't overly compelling, but are
suggestive (at least). I proposed to do a very conservative
differential experiment (very briefly: isolate two populations of
cells except for an optical link, perturb one population then look for
alterations in gene expression in the other (then use filters to
narrow the bandwidth and eventually try to induce similar changes by
providing an artificial light signal)). The terms of reference were
very clear that no pilot data was required, that the project should be
basically unfundable in a normal competition due to the speculative
nature of the hypotheses, that success should cause a significant
change in at least one field, etc., etc. Man, did I get slaughtered! I
finished dead last in the competition with the most sarcastic,
scathing review (oddly, my other review was excellent, but I still
finished dead last, so the committee was onside with the slasher). The
grants that were funded were basically those that just missed the
cutoff in the regular competition. Blech...

Damn shame, but not surprising. Even a "high risk, high reward" setup
really means just "slightly higher risk". And that's in the best
case, where there is any money to award. Here's is a little
historical anectode illustrating this (not my own experience, I just
happened to read about this).

Sometine in the late 80s, somebody in Congress (don't recall the name)
had a bright idea (happens). He reached the conclusion that the
system of research funding is rotten since it awards funds based on
specific promised results while historically the best research came
for "free play", where talented individuals try things with no
predetermined goals in mind (mirroring exactly some of the things you
wrote). So, he proposed to set aside some part of research funding,
to be granted to deserving (based on past performance) researchers
with no strings attached. Basically "here is some money, do with this
as you wish". He even managed to get this in the budget, as a pilot
program. $10M were set aside, with DOE being charged with
administering the project. Well, half a year later our congressman went to
check on DOE to see what progress was made. They informed him that
they put a significant effort into setting up the appropriate
mechanisms needed for awarding the grants. "So" he asked "how much of
the $10M did you spend so far?". "Oh, about 9.1M" was the answer.
"And, how many grants were awarded?" he asked. "Well, none yet, but
we hope to get to it soon".

So much for this one.

But, before we start berating government bureaucrats, it is important
to see what position they're in. There is a time honored rule saying
"you get what you pay for", with a corrolary to the effect that "if
you want something to happen, you must make it worthwhile for somebody
to have it happen". Now, your much maligned bureaucrat is in a
position where, he's not going to be rewarded if the affairs he manage
yield a positive result, but he can suffer if some improprity is found
out in said affairs. So, what do you think he's going to concentrate
on?


Well we certainly don't want any crap like general relativity. GPS is
nice, and all, but these "tangible benefits" are not something we want
to wait 75 years for.

A line I recall from some book goes like "If the only thing ever done
was applied research, we would've really good stone tools by now".

I saw a recent job ad inspired by fear of this kind of thing. "We want
physicists who publish in the highest impact journals" (with 2 positions
available) translates as "we are peeing ourselves in fear".

Sheesh!

1.5 weeks before closing date, I hear they had 177 applicants. I expect
they ended up with approx 200.

When I was a post-doc (mid 90's) every job I applied for had over 300
applicants (but they never insisted on publications in the highest
impact journals, as far as I remember. Then again, maybe that's why I
had so few interviews...

Heck, when I was "between jobs" at the late 80s, 400 applicants per
job wasn't rare. Getting an interview, now that was very rare.

Mati Meron | "When you argue with a fool,
meron@xxxxxxxxxxxxxxxxx | chances are he is doing just the same"
.

Relevant Pages

  • Re: Outsourcing jeopardizes U.S. chemical industry, expert says
    ... their own back yards and all they want to hear about are the success ... Are you writing grant/contract proposals for ... Care to tell us you got 100% funding success rates with ... more fish, or b. ...
    (sci.research.careers)
  • Re: Outsourcing jeopardizes U.S. chemical industry, expert says
    ... chem & eng newsgroups where they maybe never talk about careers. ... their own back yards and all they want to hear about are the success ... Are you writing grant/contract proposals for ... Care to tell us you got 100% funding success rates with ...
    (sci.research.careers)
  • Re: Teaching physics to biology students
    ... the success of their "initiatives" (My god, ... panels of scientists to rank grant applications, ... rank the impact of the applicant's recent work, and dole out funding ... "high risk, high reward" proposals. ...
    (sci.physics)
  • Re: Something ON topic - science funding
    ... is "Deficiencies in the Present Funding Process in Meteorology" ... that they were NOT having fun preparing proposals. ... "Successful grants today are more about networking and being ... nobody but Internal guys at NIH can nominate anyone for this award. ...
    (sci.research.careers)
  • Re: Something ON topic - science funding
    ... is "Deficiencies in the Present Funding Process in Meteorology" ... that they were NOT having fun preparing proposals. ... "Successful grants today are more about networking and being ... After bashing one's head against the wall ...
    (sci.research.careers)